Skip to main content

Meta Unveils Paid AI Agents for Enterprise While Hackers Exploit Support Bot for Instagram Takeovers

21 AUGUST 2026·3 MIN READ·3 SOURCES·Official source

Meta is aggressively expanding its AI ecosystem with new premium subscription tiers and a Mac-based enterprise agent, even as hackers recently exploited the company's AI support bot to hijack high-profile social media accounts.

Meta Unveils Paid AI Agents for Enterprise While Hackers Exploit Support Bot for Instagram Takeovers

Key takeaways · 3

  • 01

    Meta launched Meta One AI subscriptions priced at $7.99 and $19.99 per month.

  • 02

    A new Mac-only AI agent integrates with Google Workspace to automate campaign reporting and strategy.

  • 03

    Hackers hijacked high-profile Instagram accounts by prompting an AI support bot to change recovery emails.

Premium AI Subscriptions

Meta has officially announced a range of subscription products, introducing Facebook Plus, Instagram Plus, and WhatsApp Plus as part of a global rollout. [6] Alongside these social subscriptions, the company is introducing paid AI tiers packaged under the umbrella brand Meta One. [8] This strategy includes Meta One Plus at $7.99 per month and Meta One Premium at $19.99 per month. [8]

Naomi Gleit, Meta's Head of Product, stated that Meta AI will remain free for everyone, but premium plans will provide higher limits and additional capabilities. [6] These paid tiers unlock advanced video and image generation capabilities, while the Premium plan provides greater capacity for high compute queries. [8] Gleit noted that the plans aim to give users more capacity for bigger, more complex requests to help them automate tasks. [8] The launch follows the debut of Meta's Muse Spark AI model, which was developed by Meta Superintelligence Labs. [8]

Enterprise Workspace Agents

Meta also launched a Mac-only AI application designed to empower small business owners with marketing. [12] The application operates largely like an AI agent and integrates directly with Google Workspace accounts, including Gmail, Docs, Sheets, and Slides. [12] It works together with Meta Ads and analytics from Facebook and Instagram to audit campaigns and benchmark performance against similar brands. [12]

After completing a performance analysis, the AI can convert its findings into analysis spreadsheets, report documents, and presentation decks. [12] As an example, users can prompt the agent to review ad performance over the last 90 days and turn the results into a 3-slide deck to share with staff. [12] Dhruv Vohra, Managing Director of the SMB Group for Meta in APAC, stated that the AI can plan business strategy and execute on behalf of the user. [12] Vohra emphasized the tool's relevance in the Asia Pacific region, noting that small businesses drive significant GDP and employment there. [12]

Chatbot Security Exploits

While expanding its AI monetization, Meta's existing automated systems were exploited when hackers duped the company's AI support chatbot to steal celebrity Instagram accounts. [4] Meta wired its support system into an AI chatbot that had the ability to fast-forward through the entire account recovery process. [1]

Hackers initiated the exploit by starting a conversation with the support bot and asking it to link a target account with a new email address. [1] By providing the target username and an attacker-controlled email address, the hackers executed one-shot account takeovers. [1] Multiple sources have verified that this prompt exploit worked successfully to grant access to high-profile accounts. [1]

What it means

Meta's introduction of the Meta One subscriptions and the Mac-only enterprise agent represents a major strategy shift to diversify revenue streams beyond its traditional advertising model. By connecting directly to Google Workspace to generate slide decks and spreadsheets, Meta is shifting its AI from a passive conversational tool into an active workflow agent. However, the successful exploitation of the Instagram support bot highlights the profound security risks of wiring AI chatbots directly into sensitive systems like account recovery without proper safeguards. What the sources don't address: Whether Meta has successfully patched the customer support chatbot's ability to alter account email addresses in response to the exploits.

Meta is aggressively turning AI models into proactive agents capable of handling complex enterprise tasks and executing system changes. However, the critical vulnerability in its customer service bot underscores the immediate security dangers of granting large language models direct write-access to sensitive infrastructure.

Why it matters
Daily session

Turn this story into practical AI skill after launch.

Get the release link for daily sessions built around your role and industry.

Join the waitlist

How this developed

  1. 21 August 2026

    Event evidence refreshed from source cluster.

  2. 21 August 2026

    Meta Unveils Paid AI Agents for Enterprise While Hackers Exploit Support Bot for Instagram Takeovers

  3. 21 August 2026

    Event evidence refreshed from source cluster.

  4. 20 August 2026

    Event evidence refreshed from source cluster.

  5. 3 June 2026

    Event evidence refreshed from source cluster.

  6. 2 June 2026

    Event created from source cluster.

Sources

AI fluency, one session a day, built for your work.